wolffd@0: function engine = gaussian_inf_engine(bnet)
wolffd@0: % GAUSSIAN_INF_ENGINE Computes the joint multivariate Gaussian corresponding to the bnet
wolffd@0: % engine = gaussian_inf_engine(bnet)
wolffd@0: %
wolffd@0: % For details on how to compute the joint Gaussian from the bnet, see
wolffd@0: % - "Gaussian Influence Diagrams", R. Shachter and C. R. Kenley, Management Science, 35(5):527--550, 1989.
wolffd@0: % Once we have the Gaussian, we can apply the standard formulas for conditioning and marginalization.
wolffd@0: 
wolffd@0: assert(isequal(bnet.cnodes, 1:length(bnet.dag)));
wolffd@0: 
wolffd@0: [W, D, mu] = extract_params_from_gbn(bnet);
wolffd@0: U = inv(eye(size(W)) - W')';
wolffd@0: Sigma = U' * D * U;
wolffd@0: 
wolffd@0: engine.mu = mu;
wolffd@0: engine.Sigma = Sigma;
wolffd@0: %engine.logp = log(normal_coef(Sigma));
wolffd@0: 
wolffd@0: % This is where we will store the results between enter_evidence and marginal_nodes  
wolffd@0: engine.Hmu = [];
wolffd@0: engine.HSigma = [];
wolffd@0: engine.hnodes = [];
wolffd@0: 
wolffd@0: engine = class(engine, 'gaussian_inf_engine', inf_engine(bnet));
wolffd@0: